Lack of access to alternatives can feed distributive inferences: The view from q-spreading in children

Abstract

This paper engages with a recent analysis by Denić and Chemla (2020) who argue that what appears to be quantifier spreading in child language is in fact the result of children deriving the distributive inference. While such an analysis straightforwardly accounts for the child data involving q-spreading, they claim that a certain aspect of the adult data is left unexplained. The authors suggest a number of possible reasons for this divergence and in this paper I lay out yet another possible solution to this problem. This solution is in line with recent literature arguing that children have difficulties deriving inferences which involve alternatives obtained by lexical replacement.